Respiratory syncytial virus (RSV) is the most important respiratory pathogen of infants and young children worldwide. There is now an overwhelming body of evidence to indicate that neutralizing antibodies can mediate resistance to RSV infection and illness, including epidemiological studies which indicate that maternal antibody levels to RSV correlate with protection of infants from illness. Therefore there is a strong case that boosting levels of RSV neutralizing antibodies in very young infants at greatest risk would have marked beneficial effects. We propose to generate two highly potent neutralizing human antibodies to RSV for use as passive immunization reagents to boost infant titers either by direct administration or via administration to the mother and passive transfer to the child. The antibodies will be generated via antibody library phage display technology. We have already used this approach to yield a highly potent neutralizing antibody to RSV which is capable of drastically reducing RSV titers in the lungs of infected mice. Furthermore, for a prototype anti-HIV-1 antibody, we have shown that a human antibody derived from natural infection can be greatly improved in terms of affinity for antigen and neutralizing potency by a process of targeted mutagenesis and selection termed in vitro evolution. We propose to apply this approach to generate two highly potent evolved neutralizing human antibodies to RSV: two antibodies are preferred to one to reduce the risk of the emergence of neutralization escape mutants. The potency of the antibodies is projected to be such that protective levels achieved via maternal transmission should be maintained in the infant for more than six months post partum. The improved antibodies will further be explored as templates for vaccine design.